Add Fine-Gray weights to Fine-Gray split data — add_fg_weights_cr • ipcw

Computes the probability of remaining uncensored after a type-2 event and appends it as column p_notcens_after_death.

Usage

add_fg_weights_cr(data_long_fg, covariate = "z1", strat = "no")

Arguments

data_long_fg: A data frame in Fine-Gray format, as returned by fg_split_cr().
covariate: Character string. Name of the covariate column. Default is "z1".
strat: Character. Passed to add_ipcw_weights_cr(). "no" (default) uses a Cox model; "yes" uses stratum-specific KM estimates.

Value

data_long_fg with an additional column p_notcens_after_death.

Examples

set.seed(42)
dat <- sim_data_cr(n = 100, censoring = "baseline")
dat_long    <- wide_to_long_cr(dat)
dat_long_fg <- fg_split_cr(dat_long)
dat_long_fg <- add_fg_weights_cr(dat_long_fg, strat = "no")
summary(dat_long_fg$p_notcens_after_death)
#>      Min.   1st Qu.    Median      Mean   3rd Qu.      Max. 
#> 0.0002632 1.0000000 1.0000000 0.9525524 1.0000000 1.0000000